FourKites vs Loadsmart: Choose the Right Fit 2026

Key Takeaways

• FourKites and Loadsmart answer different jobs. FourKites sells visibility and control-tower execution. Loadsmart sells brokerage, a transportation system, and a dock, gate, and yard network.

• FourKites was founded in 2014 and still operates under that name. Loadsmart's transportation system first shipped under another name, and the dock product was an acquisition. Neither search term points at a shut-down product.

• Choose FourKites when the hole is seeing freight you do not book yourself and acting on exceptions across a wide carrier network. Choose Loadsmart when the hole is planning, pricing, tendering, docking, and settling freight, with a broker available on the same platform.

• Neither company posts a public enterprise dollar menu. Budget the subscription as quote-based, then read the vendor's own wording on what is included before you compare totals.

• Any workflow above either screen should draft the next step and wait for a person before a carrier, a customer, or a dock appointment is contacted.

• If you need both the live picture and the booking desk, you are buying two categories. Do not force this pair into a single winner.

The job this search is actually buying

Real-time transportation visibility is a live record of where a shipment is and when it should arrive, while a transportation management system is the desk that plans, tenders, and settles that freight. FourKites grew up as the first kind of company and now sells execution on top of that picture. Loadsmart grew up as a digital freight broker and now sells the second kind of company, plus the brokerage and a facility network, on one platform. The search pairs the logos. The buying decision splits on the job.

TL;DR: pick FourKites when the operations lead must see and act on freight moving through a large outside carrier network, and pick Loadsmart when the operations lead must plan, price, tender, dock, and settle freight, including loads the brokerage will haul.

FourKites was not discontinued, merged into a rival, or renamed out of the category. Matt Elenjickal founded FourKites in 2014, and the company still operates under that name, according to FourKites (2026). CB Insights lists FourKites as a Chicago company founded in 2014, a Challenger among 14 other companies in a yard-management set, and a filer of 12 patents, according to CB Insights (2026). The same page names investors that include Bain Capital Ventures, August Capital, Hyde Park Venture Partners, Mitsui, and Fabrica Ventures. Treat that Challenger label as CB Insights' market-map language, not as a score from this guide. The useful point for a buyer is simpler: the company in the search box is the same private firm founded in 2014.

Loadsmart was not discontinued either. It changed shape. In a FreightWaves interview, Loadsmart's CEO said SaaS is half of gross profit, OpenDock is in more than 4,500 facilities, the firm is a top‑70 U.S. broker, and SaaS gross profit share: 50% according to FreightWaves (2026). Read that half as the CEO's disclosure to a trade publication, not as an audited filing. It is still the clearest public evidence that the company is no longer only a digital broker. The top‑70 wording is the CEO's description in that interview. It is not a rank this guide calculated.

A brokerage operations lead, a carrier operations lead, and a shipper operations lead can all land on this query and still need different first calls. The brokerage desk often already has a system that tenders loads and wants a way to show shippers where the truck is, or it is comparing a rival broker that also sells software. The shipper desk often has a hole in procurement and execution, or a hole in arrival times on freight a third party already moves. The carrier desk is usually on the other side of someone else's mandate: a shipper may require a visibility connection, and a broker may tender freight and expect a dock appointment. Those are three relationships. Scoring the two logos on one feature grid hides the relationship you actually have.

How we evaluated these platforms

We did not run the products, and this is not a hands-on test. The evaluation is a weighted reading of public company pages, a trade interview, a company announcement carried by a news page, and a research-firm profile, all opened for this guide. Facts below are attributed to those pages. The weights are our buying logic for an operations lead who is close to a choice. They are not a measured product score.

The weight is heaviest on job match, because this query pairs two logos that do different work and a wrong category wastes the project. Network fit is next, because a visibility product is only as useful as the carriers and facilities it already sees, and a dock product is only as useful as the warehouses on it. Tender, rate, and freight audit matter for any desk that lives on acceptance and the freight bill. Facility control matters when detention starts at the door, and it matters less when the buyer does not run the warehouse. The handoff to another system matters because almost nobody gets to turn off the transportation system or the customer portal they already have. Commercial clarity matters because a quote-only enterprise deal can still be the right buy, and the buyer should know that before the second meeting. The same weighted-criteria habit shows up in other software buys, including a SaaS short list outside this category.

Add the points only for the pains you actually have. A shipper whose transportation system already tenders cleanly, and whose customers keep asking where the truck is, loads the 25-point visibility row and the 20-point network row and is already at 45 points toward a FourKites conversation. A shipper still building loads in email, with a dock calendar in a shared sheet, loads the 15-point tender row and the 10-point facility row and still has to score the visibility row before calling Loadsmart a full replacement. A brokerage that needs a capacity partner loads the tender row first. The points do not crown a winner. They stop a meeting that starts on the wrong product.

Commercial clarity is a 15-point row even though both vendors fail a public-price test. We kept the row so a buyer who will not proceed without a written number remembers to demand the quote, the volume band, and the renewal term in the same document. Dropping the row would hide a real gate. Giving either vendor full points on it would invent a price list that is not there.

What you are buying in each logo

The matrix records the public product story. It is not a test result. Where a cell says a count, that count is the company's own published figure, cited in the next section, not a number we measured.

Our reading of that grid is narrow on purpose. Overlap exists in the words both companies use for tracking, exceptions, and agents. The split that changes the contract is who accepts the tender and who runs the door. FourKites is the company you call when the freight is already someone else's to haul and your team is still chasing arrivals, dwell, and the next appointment. Loadsmart is the company you call when your team still has to turn orders into shipments, get a rate, award a carrier, meet the truck at the dock, and audit the bill. A carrier evaluating the pair is often not the buyer of either subscription. The carrier is deciding whether to accept a shipper's tracking mandate and whether to haul freight or book a door through Loadsmart's network.

If the short list on your whiteboard is two visibility networks, stop this comparison and read project44 versus FourKites for freight visibility before you put a transportation system in the same column. Mixing those conversations produces a demo that looks busy and a contract that does not cover the hole.

The counts on the public record

FourKites' about page lists a 2014 founding, 1,600+ enterprise customers, 500+ team members, six AI agents across six channels, 2.5 million+ autonomous actions a month, and Daily shipments tracked: 3.2 million according to FourKites (2026). Morningstar's reproduction of the October 14, 2026 company announcement lists 1,600+ brands, 500,000+ carriers, 176 countries, 70+ outcomes, and Network facilities: 3.6 million according to Morningstar (2026). That Morningstar page is carrying the company's announcement, and Morningstar says it does not verify third-party copy. Use the figures as the company's public network claim, with the publisher path where we read them.

Loadsmart's about page lists 1,800+ customers, 500+ employees, 100,000+ brokerage carrier relationships, a platform network of 270,000+ carriers and 4,500+ warehouses, a 2014 founding in New York, a Chicago headquarters, a November 2021 Opendock acquisition, a 2021 ShipperGuide launch under the name RFPGuide, an eight-week path from a signed AI pilot agreement to full volume, per-load brokerage quotes, a free ShipperGuide plan, quote-based managed transportation and Loadsmart AI, and Dock appointments a year: 14 million+ according to Loadsmart (2026). Keep the two carrier figures apart. The 100,000+ figure is brokerage relationships. The 270,000+ figure is the wider platform network. Adding them invents a third number the company did not publish.

The customer counts are not comparable units. FourKites' 1,600+ is enterprise brands on a visibility network. Loadsmart's 1,800+ is customers worldwide across shippers, and the company also serves warehouses and carriers on the dock product. A bigger customer count does not mean a better fit. The facility counts are even less comparable. Millions of facilities on a visibility graph means known locations the network can reason about. Thousands of warehouses on a dock network means sites where appointments, gates, and yards are actually scheduled. An operations lead who runs twelve buildings cares about the second meaning. An operations lead who buys transportation from hundreds of carriers and never operates a dock cares about the first.

The agent counts match at six only as a headcount of named programs. They are not the same work. FourKites' about page describes a digital workforce on a control tower and gives Tracy and Alan as examples of exception handling and appointment changes. Loadsmart's about page names document filing, tracking and status updates, tender failures, load audits, claims, and scheduling. Same numeral, different jobs. Do not let the matching six become a tie-breaker.

What the price pages do and do not say

Pricing checked October 14, 2026. FourKites does not post a public subscription price, so the number to put in a budget sheet is Quote-based. The public pilot terms put fees and dates in a statement of work rather than on a rate card. Loadsmart does not post a dollar subscription menu either. Use the about-page wording cited above: brokerage is priced per load with instant quotes on full truckload, the free plan on ShipperGuide is stated by the company, instant rates are described as free of a contract, and managed transportation and Loadsmart AI are quoted from volume and scope. Until a statement of work names a dollar total, treat the enterprise software and the managed service as Quote-based.

A total cost conversation still has a shape when the license line is blank. Ask each vendor to price the first year as software or service fees, implementation help, carrier or facility onboarding, and any message or appointment volume above the base. Ask who pays when a carrier must be connected one at a time. Ask what happens at renewal if your shipment count moves. Ask whether the agent work is inside the subscription or a separate quote. Write the answers in the same grid so a blank cell stays visible. A blank cell is a question for the vendor. It is not a zero.

For a brokerage, the Loadsmart per-load price is a capacity cost, similar to any other broker, and it should be compared with the brokers you already use on service and margin. It should not be compared with a FourKites subscription as if both were the same line on the invoice. For a shipper, a FourKites quote is overhead on freight you already buy elsewhere. A Loadsmart quote may replace part of that freight spend and part of the desk's software spend at the same time. Those savings are not automatic. They show up only if the lanes you move are lanes they can cover, and only if your team will actually award freight through the new desk.

The eight-week pilot length on Loadsmart's about page is a planning input, not a promise that your network will be live on that day. Ask what "full volume" means for your lane count, which modes are inside the pilot, and who closes the exceptions the agents leave behind. FourKites does not publish a standard week count for a customer go-live on the pages reviewed here. Ask which of the six agents are in the first release, which modes are included, and which messages an agent may send before a person sees them.

FourKites when the tower has to see the freight

FourKites fits a shipper, a third-party logistics provider, or a brokerage control tower that already has a system of record for planning and tendering and is still blind between pickup and delivery. The public story runs from road, rail, ocean, and air coverage built in the early years to a control tower that holds digital twins of orders, shipments, inventory, and assets. The about page, cited above, describes three acquisitions that widened that graph: Haven for ocean, NIC-place for a European multimodal network, and TrackX for yard and facility data. Offices listed there are Chicago, Amsterdam, Munich, Chennai, and Singapore. The commercial model on that page is Outcome-as-a-Service, which the October 14, 2026 announcement frames as work organized around results the buyer already reports, such as on-time performance, rather than around seats.

Best fit is the team whose painful hour is exception handling on freight other companies haul. The network claim that matters is the carrier and facility graph, because a predictive arrival is only useful if the underlying movement data shows up without a months-long connection project on every lane. Limitations follow from the category. FourKites is not the broker who accepts your tender and sends a truck. It is not a dock-scheduling network you roll out to every warehouse you ship to. A yard acquisition is not the same asset as a scheduling product used by thousands of buildings. If your team still builds loads in spreadsheets, a control tower has nothing solid to watch.

Implementation, on the public record, is a statement of work. A practical first release is one mode, one region, and a short list of exception types a person still approves. Ask for the field list your transportation system must send, the carrier connections included in the base network, and the audit record an agent leaves when it changes an appointment. The company describes Tracy resolving exceptions and Alan rescheduling appointments. Treat those names as product examples from the vendor, not as staff you can hire, and put a human review on any customer-facing change until your own team trusts the rule.

Primary evidence is the company about page and the October 14, 2026 announcement cited above, plus the CB Insights profile for the 2014 founding and the private-company status. That set is enough to decide whether FourKites belongs on the short list. It is not enough to skip a scoped pilot on your own lanes.

Loadsmart when the desk has to move the freight

Loadsmart fits a shipper or a manufacturer that wants one platform for planning, procurement, execution, and settlement, and that is willing to book at least some freight with the company's brokerage. It also fits a warehouse team that needs dock, gate, and yard scheduling on Opendock, including sites that are not ready to replace a transportation system. The about page is explicit that existing systems can stay: connections are described over API, EDI, and MCP, with outcomes written back to the system of record. ShipperGuide can also be the transportation system. Brokerage is available and is not required for every load. That optionality is the point a buyer should test. Ask which workflows quietly assume you will tender to Loadsmart when your own carrier rejects a load.

Best fit is a mid-market or enterprise shipper whose planners still touch every load, whose dock calendar is outside the transportation system, and whose audit team rebuilds the freight bill by hand. The six agents named on the about page map to that pain: documents, status updates, tender failures, load audits, claims, and scheduling. Limitations are the mirror image of FourKites. Loadsmart does not publish a daily shipment-tracking volume comparable to a visibility network, and its country coverage is described as North America, South America, and EMEA rather than as a stated country count. A brokerage relationship count and a platform carrier count are different pools. A shipper that needs historical lane behavior across a carrier base it will not move onto this platform should not expect the visibility graph of a company that has spent a decade collecting other companies' freight events.

Founders Felipe Capella and Ricardo Salgado started the company in New York in 2014. Investors named on the about page include BlackRock, Maersk Growth, Ports America, TFI International, Chromo Invest, and Connor Capital. The company states that a SOC 2 Type II audit is complete. Those facts help a security review start. They do not replace your own questionnaire.

Picture a Monday file of creates moving through code instead of the shipment screen: the public limit page sets a default of 5 requests per minute, tells the caller to back off on status 429, and states Shipment creates allowed: 100 per minute according to Loadsmart (2026), so 100 divided by 5 is 20 and the create route is 20 times the default, while 100 calls a minute across 60 minutes is 6,000 creates and the default path across that same hour is 300 calls (5 times 60), and a charge or milestone later in the day posts against the same shipment_id.

That table is Loadsmart's published cap, multiplied by 60 minutes, not a volume we observed on a customer account. A desk that bursts past the cap will spend the hour retrying status 429 instead of moving freight. A desk that stays under the cap still needs a person on rejects, claims, and any appointment change a customer will feel. The identifier to store in your own log is shipment_id, because later charges, cancellations, and milestones on that same page hang off it.

Who this is for

This comparison is for an operations lead at a freight brokerage, a trucking carrier, or a shipper who has already decided that spreadsheets and inboxes are failing, and who typed the two names because a colleague, a shipper mandate, or a demo queue put both logos on the list. It is for the person who can say, in one sentence, whether the pain is "we cannot see the freight" or "we cannot plan, buy, dock, and pay the freight." It is also for the person who will sit in procurement and explain why a visibility subscription and a brokerage invoice are different purchases.

You are in the right place if you can name the system that stays. A shipper keeping an existing transportation system, and adding eyes on the carrier base, is a FourKites conversation with a possible Loadsmart conversation only on lanes they want the brokerage to haul. A shipper replacing the desk that builds loads is a Loadsmart conversation, and a FourKites conversation only if a separate network must watch carriers Loadsmart will not run. A carrier is in the right place if a customer is pushing a tracking connection or a dock-appointment rule, and the decision is whether to comply, not whether to buy a control tower for the fleet.

Red flags: you need a broker to accept the tender and send a truck, and you are scoring FourKites as that broker; you need a multi-enterprise tracking graph for carriers you will keep tendering yourself, and you are scoring Loadsmart's brokerage as that graph; you will not sign without a public enterprise rate card, because neither vendor posts one.

A different operations desk, in-house membership renewals, has the same shape of problem: automate the repeatable step, and keep a person on the exception. The freight version of that rule is identical. The vendor can recommend a status change. Your team still owns the customer promise.

Where a review step sits above the screen

The painful hour is rarely the tracking ping. It is the moment a late arrival, a rejected tender, or a missed door has to become a message to a carrier or a change to an appointment, and the person who should approve it is in another meeting. Both vendors describe agents that can do parts of that work inside their own products. Many desks still have a second system, a customer portal, or an inbox that those agents do not touch. That gap is where a configurable layer earns its keep, and where it becomes clutter if the vendor screen was already enough.

When a late arrival or a rejected tender has to leave the vendor screen, a proposed US Tech Automations workflow can take the export the buyer is already licensed to pull, match the load to a lane rule the operations lead configured, and place a draft carrier note in a review queue. The trigger is the export or the webhook the source system already provides. The action is the match against the rule. The output is the queued draft, not a sent email.

In that configurable design, US Tech Automations would not send the note on its own. The prerequisite is a stable shipment identifier in the file or payload, credentials the buyer stores and can revoke, and a written map from the source fields to the draft. A person approves the outbound message, or the item stays held. The output of the approval is an audit row that shows the trigger, the draft, and the decision. Nothing in that design is a live customer result. It is a workflow you would specify, review, and turn on only after the export proves it carries the identifier every time.

The no-code path, fairly compared

The real alternative for many desks is not another freight suite. It is a chain built in Zapier, Make, or n8n, or a small service the internal team maintains. Those tools can support run histories, retries, error branches, and audit evidence when someone configures them. A careful builder can store each attempt, branch when a call fails, and keep a log a manager can open. That is real operational machinery, and dismissing it as toys leads buyers to overbuy.

The buyer who chooses that path must design and own observability, idempotency, escalation, access control, and maintenance. Observability means someone notices when the chain stops before a customer does. Idempotency means a repeated event does not create a second tender or a second appointment change. Escalation means a stuck load reaches a named person during the shift, not the next morning. Access control means the credential that can read freight cannot also email every carrier. Maintenance means a field rename in the source system is someone's job next quarter.

A proposed US Tech Automations design can set the idempotency key to the source shipment id, branch when the same event arrives twice, cap retries so a loop cannot hammer a rate limit, and stop before any carrier email or dock change, with the same prerequisites: a licensed export, a stable id, and a person on the send. The difference to specify is who watches the chain and who is on the hook when a rule is wrong. If your team already has that ownership in Zapier, Make, or n8n, staying there is a rational choice. If the chain has become a pile of private zaps nobody can explain, a configured review queue with an audit row is the conversation to have.

When NOT to use US Tech Automations. A simpler path wins when the team already works inside one vendor login and the exception never has to notify a carrier, a customer, or another system. It also wins when a single automation inside the transportation system covers the lane and a person already sees the failure in that inbox. It wins again when the only gap is a missing field on a document the vendor screen already stores, because a second product would create a second place to check.

Mistakes that waste the short list

Treat these as disqualifiers for the process, not as insults to either vendor.

• Scoring the two products as one category because both say "tracking" and "AI." The contract you sign follows the job, not the adjective.

• Pasting a generic transportation-software price range from a blog into the budget. If it is not on that vendor's own pricing page, it is not their price. Both enterprise offers here are quote-based.

• Adding Loadsmart's brokerage relationship count to the platform carrier count, or comparing that sum with FourKites' carrier graph. The labels are different, and the sum is not a published figure.

• Letting an agent message a carrier or move a door with no human hold, then discovering the audit trail after a customer complains.

• Writing a create loop that ignores the published per-minute cap and then blaming the platform for a backlog of retries.

• Starting implementation before you name the system that remains the system of record. Two systems of record means two ETAs, and the customer will hear both.

A short decision checklist keeps the next meeting honest. Name the job in one sentence, see freight or book and settle freight. Match that sentence to a category before any feature demo. Read the public counts and keep every carrier figure labeled. Ask for the quote, the volume band, the modes in the first release, and the agent actions that require a person. Decide who may approve a carrier message. Only then look at a layer above the system you intend to keep.

Terms worth agreeing on

Agree on these words in the statement of work. Vendors use them in overlapping ways. Your team will use them in arguments about who was late. A shared glossary will not settle a claim by itself. It will stop a demo in which both companies say yes to "tracking" and mean different work.

Questions before you book the call

Are FourKites and Loadsmart substitutes?

No. FourKites is a visibility and control-tower company, and Loadsmart is a brokerage plus a transportation system and a dock, gate, and yard network. They share tracking language and they do not share the same contract.

Did either product get renamed, merged, or shut down?

No. FourKites was founded in 2014 and still operates as FourKites. Loadsmart still operates, its transportation system first shipped under another name, and its dock product came in through an acquisition. Use the sourced history above for the founding year and for Loadsmart's earlier product name.

Is there a public price list for either vendor?

No. Both enterprise offers are quote-based on the pages reviewed for this guide. Loadsmart additionally states a free software plan and per-load brokerage pricing, and those details still are not a full enterprise rate card.

Which desk should open FourKites first?

Open FourKites first when a transportation system already plans and tenders the freight and the remaining pain is arrival time, exceptions, and a network view across carriers you will keep. Open Loadsmart first when the pain is building, pricing, awarding, docking, and paying the load.

Can a no-code chain replace either platform?

No. Zapier, Make, or n8n can move a status between systems you already have, and they do not supply a carrier visibility graph or a broker's truck. Use them for the handoff, not as a substitute for the category you still lack.

What must be true before any outside workflow messages a carrier?

A person has to approve the message. The export also has to carry a stable shipment id, the buyer has to be licensed to pull the feed, and a repeated event must not send the message twice.

What should the carrier's operations lead do with this pair?

Comply with a shipper's tracking mandate, or book a door and haul tendered freight, only after you know which company is asking and what data they require. Buying a control tower for your own fleet is a separate project from accepting someone else's connection rule.

The operations lead who can answer those seven questions in a single sitting is ready for a vendor call. The lead who cannot is ready for an internal meeting that names the job, the system that stays, and the person who may approve a carrier note. If that meeting shows the real gap is the handoff between a system you will keep and a person who must sign off, see how US Tech Automations configures this as a review step above that system, with the export, the hold, and the audit row specified before anything is turned on.